Error Encountered Trying to Load Images when Initializing Cloudera Data Science Workbench

Here are some sample error messages you might see when initializing Cloudera Data Science Workbench or adding a Worker node.

Error encountered while trying to load images.: 1
Unable to load images from [/etc/cdsw/images/cdsw_<version>.tar.gz].: 1
Error processing tar file(exit status 1): write /../..tar: no space left on device

These errors are an indication that the root volume hosting the docker-tmp is running out of space when trying to initialize Cloudera Data Science Workbench or start a Worker role. During the initialization process, the Cloudera Data Science Workbench installer temporarily decompresses the Legacy Engine and ML Runtime image files located in the CDSW parcel to the /var/lib/cdsw/docker-tmp/ directory.


  • If you have previously partitioned the root volume (which should be at least 200 GB), make sure you allocate at least 200 GB to /var/lib/cdsw/docker-tmp/ so that the installer can proceed without running out of space.
  • For every host running out of space, update the environment variable $DOCKER_TMP inside the file:
    to point to any location on the host that has 200 GB or more, then rerun whichever steps caused this issue in the first place.